Hugging Face's logo

chuanmew
/
aubo

Safetensors
Model card Files
xet
 Community
chuanmew commited on
Commit
3665bc9
·
verified ·
1 Parent(s): 998cf9f

test-4-checkpoint-500

Browse files
Files changed (14) hide show
  1. test-4-checkpoint-500/config.json +133 -0
  2. test-4-checkpoint-500/dataset_statistics.json +253 -0
  3. test-4-checkpoint-500/model.safetensors +3 -0
  4. test-4-checkpoint-500/optimizer.pt +3 -0
  5. test-4-checkpoint-500/rng_state_0.pth +3 -0
  6. test-4-checkpoint-500/rng_state_1.pth +3 -0
  7. test-4-checkpoint-500/rng_state_2.pth +3 -0
  8. test-4-checkpoint-500/rng_state_3.pth +3 -0
  9. test-4-checkpoint-500/rng_state_4.pth +3 -0
  10. test-4-checkpoint-500/rng_state_5.pth +3 -0
  11. test-4-checkpoint-500/rng_state_6.pth +3 -0
  12. test-4-checkpoint-500/rng_state_7.pth +3 -0
  13. test-4-checkpoint-500/scheduler.pt +3 -0
  14. test-4-checkpoint-500/trainer_state.json +68 -0
test-4-checkpoint-500/config.json ADDED
@@ -0,0 +1,133 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "batch_size": 64,
3
+ "buffer_size": 64,
4
+ "from_pretrained": "oxe-g2-checkpoint-300000",
5
+ "frozen_keys": [
6
+ "*hf_model*"
7
+ ],
8
+ "model": {
9
+ "heads": {
10
+ "action": {
11
+ "args": [],
12
+ "kwargs": {
13
+ "action_dim": 7,
14
+ "action_horizon": 4,
15
+ "dropout_rate": 0.0,
16
+ "n_diffusion_samples": 1,
17
+ "readout_key": "readout_action",
18
+ "token_embedding_size": 768,
19
+ "use_map": false
20
+ },
21
+ "module": "octo.model.components.action_heads",
22
+ "name": "DiffusionActionHead"
23
+ }
24
+ },
25
+ "max_horizon": 10,
26
+ "observation_tokenizers": {
27
+ "primary": {
28
+ "args": [],
29
+ "kwargs": {
30
+ "encoder": {
31
+ "args": [],
32
+ "kwargs": {
33
+ "in_features": 6
34
+ },
35
+ "module": "octo.model.components.vit_encoders",
36
+ "name": "SmallStem16"
37
+ },
38
+ "obs_stack_keys": [
39
+ "image_primary"
40
+ ],
41
+ "task_stack_keys": [
42
+ "image_primary"
43
+ ]
44
+ },
45
+ "module": "octo.model.components.tokenizers",
46
+ "name": "ImageTokenizer"
47
+ }
48
+ },
49
+ "readouts": {
50
+ "action": 1
51
+ },
52
+ "repeat_task_tokens": true,
53
+ "task_tokenizers": {
54
+ "language": {
55
+ "args": [],
56
+ "kwargs": {
57
+ "encoder": "t5-base",
58
+ "finetune_encoder": false
59
+ },
60
+ "module": "octo.model.components.tokenizers",
61
+ "name": "LanguageTokenizer"
62
+ }
63
+ },
64
+ "token_embedding_size": 768,
65
+ "transformer_kwargs": {
66
+ "add_position_embedding": false,
67
+ "attention_dropout_rate": 0.0,
68
+ "dropout_rate": 0.0,
69
+ "mlp_dim": 3072,
70
+ "num_attention_heads": 12,
71
+ "num_layers": 12
72
+ },
73
+ "use_correct_attention": true
74
+ },
75
+ "normalize_method": "sign",
76
+ "obs_token_nums": {
77
+ "primary": 256
78
+ },
79
+ "sampler_num_samples": 2560000,
80
+ "seed": 2,
81
+ "subsample_length": 99999,
82
+ "text_processor": {
83
+ "args": [],
84
+ "kwargs": {
85
+ "encode_with_model": false,
86
+ "tokenizer_kwargs": {
87
+ "max_length": 16,
88
+ "padding": "max_length",
89
+ "return_tensors": "np",
90
+ "truncation": true
91
+ },
92
+ "tokenizer_name": "t5-base"
93
+ },
94
+ "module": "octo.components.text_processing",
95
+ "name": "HFTokenizer"
96
+ },
97
+ "training_arguments": {
98
+ "bf16": true,
99
+ "dataloader_pin_memory": true,
100
+ "ddp_find_unused_parameters": true,
101
+ "eval_steps": 5000,
102
+ "gradient_accumulation_steps": 4,
103
+ "learning_rate": 5e-05,
104
+ "logging_nan_inf_filter": false,
105
+ "logging_steps": 100,
106
+ "lr_scheduler_kwargs": {
107
+ "decay_type": "cosine",
108
+ "min_lr_ratio": 0.05,
109
+ "num_decay_steps": 298000
110
+ },
111
+ "lr_scheduler_type": "constant",
112
+ "max_grad_norm": 1.0,
113
+ "max_steps": 300000,
114
+ "optim": "adamw_torch_fused",
115
+ "per_device_eval_batch_size": 1,
116
+ "per_device_train_batch_size": 1,
117
+ "report_to": "wandb",
118
+ "save_safetensors": true,
119
+ "save_steps": 500,
120
+ "torch_compile": true,
121
+ "warmup_steps": 2000,
122
+ "weight_decay": 0.1
123
+ },
124
+ "training_keys": [
125
+ "*action*",
126
+ "*task_projections*",
127
+ "*obs_projections*",
128
+ "*readout_embeddings*",
129
+ "*task_pos_embeddings*",
130
+ "*obs_pos_embeddings*",
131
+ "*observation_tokenizers*"
132
+ ]
133
+ }
test-4-checkpoint-500/dataset_statistics.json ADDED
@@ -0,0 +1,253 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "episode_index": {
4
+ "min": [
5
+ 0
6
+ ],
7
+ "max": [
8
+ 128
9
+ ],
10
+ "mean": [
11
+ 63.14946338383838
12
+ ],
13
+ "std": [
14
+ 35.1467891774593
15
+ ],
16
+ "count": [
17
+ 12672
18
+ ]
19
+ },
20
+ "task_index": {
21
+ "min": [
22
+ 0
23
+ ],
24
+ "max": [
25
+ 3
26
+ ],
27
+ "mean": [
28
+ 1.6840277777777777
29
+ ],
30
+ "std": [
31
+ 1.1113384592218372
32
+ ],
33
+ "count": [
34
+ 12672
35
+ ]
36
+ },
37
+ "observation.state": {
38
+ "min": [
39
+ -0.6913691163063049,
40
+ -0.25204867124557495,
41
+ -0.22021731734275818,
42
+ -3.140528440475464,
43
+ -1.213579535484314,
44
+ -3.1340115070343018,
45
+ 0.0
46
+ ],
47
+ "max": [
48
+ -0.26134854555130005,
49
+ 0.40225377678871155,
50
+ 0.4610653519630432,
51
+ 3.1412761211395264,
52
+ 0.644866406917572,
53
+ 3.1401240825653076,
54
+ 1.0
55
+ ],
56
+ "mean": [
57
+ -0.4765559839921994,
58
+ 0.03277433536791847,
59
+ 0.01813845893630427,
60
+ 1.4840563985447381,
61
+ -0.1872646863000659,
62
+ -1.201557804914303,
63
+ 0.13328598546378098
64
+ ],
65
+ "std": [
66
+ 0.09873974979292631,
67
+ 0.12291375325367411,
68
+ 0.13182701724212956,
69
+ 2.530463134923631,
70
+ 0.39532953579234853,
71
+ 1.3639962652103377,
72
+ 0.33988356830665484
73
+ ],
74
+ "count": [
75
+ 12672
76
+ ]
77
+ },
78
+ "frame_index": {
79
+ "min": [
80
+ 0
81
+ ],
82
+ "max": [
83
+ 225
84
+ ],
85
+ "mean": [
86
+ 55.168955176767675
87
+ ],
88
+ "std": [
89
+ 39.5185866669507
90
+ ],
91
+ "count": [
92
+ 12672
93
+ ]
94
+ },
95
+ "index": {
96
+ "min": [
97
+ 0
98
+ ],
99
+ "max": [
100
+ 12671
101
+ ],
102
+ "mean": [
103
+ 6335.5
104
+ ],
105
+ "std": [
106
+ 3658.0912941951938
107
+ ],
108
+ "count": [
109
+ 12672
110
+ ]
111
+ },
112
+ "action": {
113
+ "min": [
114
+ -0.009999999776482582,
115
+ -0.009999999776482582,
116
+ -0.009999999776482582,
117
+ -0.06283185631036758,
118
+ -0.06283185631036758,
119
+ -0.06283185631036758,
120
+ 0.0
121
+ ],
122
+ "max": [
123
+ 0.009999999776482582,
124
+ 0.009999999776482582,
125
+ 0.009999999776482582,
126
+ 0.06283185631036758,
127
+ 0.06283185631036758,
128
+ 0.06283185631036758,
129
+ 1.0
130
+ ],
131
+ "mean": [
132
+ -0.0010779668760702325,
133
+ -0.000870422678087375,
134
+ -0.0009524932320082003,
135
+ -0.00011404143052376722,
136
+ 0.001646163140721359,
137
+ 0.0006842485724881945,
138
+ 0.14094065714126566
139
+ ],
140
+ "std": [
141
+ 0.0050928476793778805,
142
+ 0.0052109778002382215,
143
+ 0.0062727072406226275,
144
+ 0.01305372613557191,
145
+ 0.014969692486291531,
146
+ 0.021519857205295197,
147
+ 0.3479603303557627
148
+ ],
149
+ "count": [
150
+ 12672
151
+ ],
152
+ "mask": [
153
+ true,
154
+ true,
155
+ true,
156
+ true,
157
+ true,
158
+ true,
159
+ false
160
+ ]
161
+ },
162
+ "timestamp": {
163
+ "min": [
164
+ 0.0
165
+ ],
166
+ "max": [
167
+ 22.5
168
+ ],
169
+ "mean": [
170
+ 5.516895517676768
171
+ ],
172
+ "std": [
173
+ 3.9518586666950704
174
+ ],
175
+ "count": [
176
+ 12672
177
+ ]
178
+ },
179
+ "observation.images.main_camera": {
180
+ "min": [
181
+ [
182
+ [
183
+ 0.0
184
+ ]
185
+ ],
186
+ [
187
+ [
188
+ 0.0
189
+ ]
190
+ ],
191
+ [
192
+ [
193
+ 0.0
194
+ ]
195
+ ]
196
+ ],
197
+ "max": [
198
+ [
199
+ [
200
+ 1.0
201
+ ]
202
+ ],
203
+ [
204
+ [
205
+ 1.0
206
+ ]
207
+ ],
208
+ [
209
+ [
210
+ 1.0
211
+ ]
212
+ ]
213
+ ],
214
+ "mean": [
215
+ [
216
+ [
217
+ 0.56009308397514
218
+ ]
219
+ ],
220
+ [
221
+ [
222
+ 0.5088963212208076
223
+ ]
224
+ ],
225
+ [
226
+ [
227
+ 0.47626117961949455
228
+ ]
229
+ ]
230
+ ],
231
+ "std": [
232
+ [
233
+ [
234
+ 0.27688289864943944
235
+ ]
236
+ ],
237
+ [
238
+ [
239
+ 0.26883972589038224
240
+ ]
241
+ ],
242
+ [
243
+ [
244
+ 0.29031452356480364
245
+ ]
246
+ ]
247
+ ],
248
+ "count": [
249
+ 11006
250
+ ]
251
+ }
252
+ }
253
+ ]
test-4-checkpoint-500/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fda91e5a4ce79eccd1acc1b4649683fdf4edcaa1164b5d71a319bc355b02c3b5
3
+ size 401078544
test-4-checkpoint-500/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a01d89421b32bda5e882530b8596625e54b699943e05efdd01ef1732f16f8171
3
+ size 26356666
test-4-checkpoint-500/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0d72845b2daf8aed25b47f8eb106c0ed64de5b2d41edcc7299d9509c85934a67
3
+ size 15984
test-4-checkpoint-500/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8d941af5f9358a5746c30b26fddf4704952810ef9beeb53d6361e278bd67855d
3
+ size 15984
test-4-checkpoint-500/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a7aa380c168aa3822dff7015cb5cbda935e9d36a2535a072e3acff4d5f100bba
3
+ size 15984
test-4-checkpoint-500/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7bfb72b1bc310cd0eee525f0a5e9455ce83a847ac714dd0400c988e72778a412
3
+ size 15984
test-4-checkpoint-500/rng_state_4.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:22d7b9f2a97beede4dfb45283ee46b929dead43898c3bd4b5c49aaedfe4114f0
3
+ size 15984
test-4-checkpoint-500/rng_state_5.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:51610dd9296a2f27da4ed26bad68f3a3c0e160bb9c44abe5f51cf7f2098ec1cf
3
+ size 15984
test-4-checkpoint-500/rng_state_6.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f2e94dac679100a6194ec96fc28adbedc91e577f1d11bb853ccb337aa50311e5
3
+ size 15984
test-4-checkpoint-500/rng_state_7.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b10be1f55437e630703942aa11a2078fb1ff717515fce680a97af6c5137e29d7
3
+ size 15984
test-4-checkpoint-500/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:722095417950d5a40eae3f57d8e53a42def012f8bda7ae218d2ae2f6092e373c
3
+ size 1064
test-4-checkpoint-500/trainer_state.json ADDED
@@ -0,0 +1,68 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.4,
5
+ "eval_steps": 5000,
6
+ "global_step": 500,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.08,
13
+ "grad_norm": 3.25,
14
+ "learning_rate": 5e-05,
15
+ "loss": 0.5172,
16
+ "step": 100
17
+ },
18
+ {
19
+ "epoch": 0.16,
20
+ "grad_norm": 3.921875,
21
+ "learning_rate": 5e-05,
22
+ "loss": 0.3445,
23
+ "step": 200
24
+ },
25
+ {
26
+ "epoch": 0.24,
27
+ "grad_norm": 4.09375,
28
+ "learning_rate": 5e-05,
29
+ "loss": 0.3051,
30
+ "step": 300
31
+ },
32
+ {
33
+ "epoch": 0.32,
34
+ "grad_norm": 3.09375,
35
+ "learning_rate": 5e-05,
36
+ "loss": 0.2824,
37
+ "step": 400
38
+ },
39
+ {
40
+ "epoch": 0.4,
41
+ "grad_norm": 3.09375,
42
+ "learning_rate": 5e-05,
43
+ "loss": 0.265,
44
+ "step": 500
45
+ }
46
+ ],
47
+ "logging_steps": 100,
48
+ "max_steps": 300000,
49
+ "num_input_tokens_seen": 0,
50
+ "num_train_epochs": 240,
51
+ "save_steps": 500,
52
+ "stateful_callbacks": {
53
+ "TrainerControl": {
54
+ "args": {
55
+ "should_epoch_stop": false,
56
+ "should_evaluate": false,
57
+ "should_log": false,
58
+ "should_save": true,
59
+ "should_training_stop": false
60
+ },
61
+ "attributes": {}
62
+ }
63
+ },
64
+ "total_flos": 0.0,
65
+ "train_batch_size": 1,
66
+ "trial_name": null,
67
+ "trial_params": null
68
+ }